com.imsl.datamining.neural
Class QuasiNewtonTrainer.Objective
java.lang.Object
com.imsl.datamining.neural.QuasiNewtonTrainer.Objective
- All Implemented Interfaces:
- MinUnconMultiVar.Function
- Direct Known Subclasses:
- QuasiNewtonTrainer.BlockObjective, QuasiNewtonTrainer.GradObjective
- Enclosing class:
- QuasiNewtonTrainer
- protected class QuasiNewtonTrainer.Objective
- extends Object
- implements MinUnconMultiVar.Function
The Objective class is passed to the optimizer.
Method Summary |
double |
f(double[] weights)
Public interface for the multivariate function to be minimized. |
Methods inherited from class java.lang.Object |
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait |
nFunctionEvaluations
protected int nFunctionEvaluations
nObs
protected int nObs
nY
protected int nY
f
public double f(double[] weights)
- Description copied from interface:
MinUnconMultiVar.Function
- Public interface for the multivariate function to be minimized.
- Specified by:
f
in interface MinUnconMultiVar.Function
- Parameters:
weights
- a double
array, the point
at which the function is to be evaluated
- Returns:
- a
double
, the value of the function at x
Copyright 1970-2006 Visual Numerics, Inc.
Built June 1 2006.